I realized that the HLL Riemann solver does not correctly capture shear waves, so I've now implemented an HLLC like solver that has a constant density as well as pressure and velocity in the mid-state, but allows the transverse velocities to vary. This option is only switched on when the simulation is 'isothermal'. It should work for both the isothermal and locally isothermal EoS.
While the solver is idealistically better, in some simple tests I can't see any difference...
I realized that the HLL Riemann solver does not correctly capture shear waves, so I've now implemented an HLLC like solver that has a constant density as well as pressure and velocity in the mid-state, but allows the transverse velocities to vary. This option is only switched on when the simulation is 'isothermal'. It should work for both the isothermal and locally isothermal EoS.
While the solver is idealistically better, in some simple tests I can't see any difference...